Manoeuvres aimed at increasing beta cell mass have been proposed as regenerative medicine strategies for diabetes treatment. Raf-1 kinase inhibitor protein 1 (RKIP1) is a common regulatory node of the mitogen-activated protein kinase (MAPK) and nuclear factor κB (NF-κB) pathways and therefore may be involved in regulation of beta cell homeostasis. The aim of this study was to investigate the involvement of RKIP1 in the control of beta cell mass and function.
Rkip1 (also known as Pebp1) knockout (Rkip1 −/−) mice were characterised in terms of pancreatic and glucose homeostasis, including morphological and functional analysis. Glucose tolerance and insulin sensitivity were examined, followed by assessment of glucose-induced insulin secretion in isolated islets and beta cell mass quantification through morphometry. Further characterisation included determination of endocrine and exocrine proliferation, apoptosis, MAPK activation and whole genome gene expression assays. Capacity to reverse a diabetic phenotype was assessed in adult Rkip1 −/− mice after streptozotocin treatment.
Rkip1 −/− mice exhibit a moderately larger pancreas and increased beta cell mass and pancreatic insulin content, which correlate with an overall improvement in whole body glucose tolerance. This phenotype is established in young postnatal stages and involves enhanced cellular proliferation without significant alterations in cell death. Importantly, adult Rkip1 −/− mice exhibit rapid reversal of streptozotocin-induced diabetes compared with control mice.
These data implicate RKIP1 in the regulation of pancreatic growth and beta cell expansion, thus revealing RKIP1 as a potential pharmacological target to promote beta cell regeneration.
KeywordsBeta cell Diabetes Insulin Neonatal beta cell growth Proliferation Raf-1 RKIP
Extracellular signal-regulated kinase
Glycogen synthase kinase 3
Intraperitoneal glucose tolerance test
Mitogen-activated protein kinase
MAP kinase kinase
Nuclear factor κB
Platelet-derived growth factor
Raf-1 kinase inhibitor protein 1
Type 2 diabetes results from pancreatic beta cell failure on a background of insulin resistance. To compensate for the increased insulin demand in insulin-resistant states, beta cells augment their insulin biosynthetic and secretory capacity in part by expanding their mass through hyperplastic and hypertrophic mechanisms . However, when beta cells reach their maximum compensatory response and can no longer overcome insulin resistance, overt diabetes occurs. Hence, understanding how beta cells regulate their growth could lead to the development of therapeutic approaches to prevent and/or delay the onset of this disease.
During the last few years, our group has been investigating the actions of sodium tungstate on beta cell plasticity. We have shown that this agent can normalise blood glucose levels in streptozotocin (STZ)-treated rats and that this normalisation correlates with increased beta cell proliferation and neogenesis [2–4]. Recently, in a global transcriptome analysis, we identified the gene encoding the Raf-1 kinase inhibitor protein 1 (RKIP1) as one of the potential candidate genes involved in the ability of tungstate to modulate pancreatic beta cell mass .
RKIP1 is an evolutionarily conserved protein that was first recognised as a regulator of the MAPK signal transduction cascade through its interaction and inhibition of Raf-1 kinase activity . However, upon phosphorylation by protein kinase C, RKIP1 can dissociate from Raf-1 and inhibit G protein-coupled receptor kinase 2 (GRK2) , a negative regulator of G protein-coupled receptors. Additionally, RKIP1 has been implicated as a negative regulator of nuclear factor κB (NF-κB)-signalling pathways and an activator of glycogen synthase kinase 3β (GSK3β)-dependent signalling pathways [8, 9]. Therefore, RKIP1 mediates the crosstalk between various important cellular signalling pathways and, consequently, can influence a wide array of cellular functions including cell proliferation, differentiation and apoptosis. Thus, RKIP1 has been implicated in cancer, neurological functioning and reproduction (reviewed in [10–12]).
In the pancreas, RKIP1 is present at high levels in islets, whilst its presence in exocrine cells is somewhat controversial [13, 14]. The role of RKIP1 in beta cells is not fully understood. Because modulation of RKIP1 levels in the beta cell line HIT-T15 influences MAPK signalling and cell proliferation, it has been speculated that the presence of high amounts of RKIP1 in adult beta cells could partially explain the low proliferative capacity of these cells . In line with this, overproduction of Raf-1 kinase in primary beta cells enhances cell proliferation . Conversely, Raf-1 kinase inhibition results in increased beta cell apoptosis, indicating that modulation of this pathway may also contribute to beta cell survival . Taken together, this in vitro evidence supports a possible role of RKIP1 in the modulation of pancreatic beta cell plasticity.
The aim of the present study was to investigate the involvement of RKIP1 in the regulation of beta cell mass using mice with a homozygous germ-line deletion of the Rkip1 gene (Rkip1 −/−). The phenotypes of Rkip1 −/− mice described to date include age-dependent learning and olfaction deficits and fertility defects within the males [12, 17]. Here we have characterised pancreatic morphology and glucose homeostasis in Rkip1 −/− and provide evidence of a role of RKIP1 as a negative regulator of beta cell mass under physiological (normal growth) and pathological (STZ-induced diabetes) conditions in vivo.
Rkip1 −/− mice were generously provided by Dr Jan Klysik (Brown University, Providence, Rhode Island, USA), maintained in a C57BL/6 background and genotyped as described previously . Protocols were approved by the Animal Ethics Committee of the University of Barcelona and the Principles of Laboratory Animal Care were followed. Diabetes was induced by intraperitoneal injection of 100 mg/kg STZ (Sigma-Aldrich, St Louis, MO, USA) in adult mice weighing 18–22 g that were fasted for 6 h. Blood glucose levels were measured weekly until the time they were killed.
For intraperitoneal glucose tolerance tests (IPGTTs), mice fasted for 6 h were injected with 2 g glucose/kg body weight. Blood samples were collected at 0, 15, 30, 60 and 120 min after injection and glycaemia was measured using a clinical glucometer and Accu-Chek test strips (Roche, Basel, Switzerland). Plasma was separated by centrifugation and kept at −80°C for insulin determination. For insulin tolerance tests, animals were injected with 0.75 U insulin/kg body weight. Blood samples were collected at 0, 15, 30 and 60 min and glucose levels were measured as indicated above.
Islets were isolated from 28-day-old (P28) or 8-week-old male mice by collagenase digestion. P28 islets were individually handpicked under a stereomicroscope after undergoing collagenase digestion, whereas adult islets were purified using a Histopaque gradient (Sigma-Aldrich) . Insulin secretion was evaluated using freshly isolated islets in static incubation assays . Insulin was measured by ELISA (Mercodia, Uppsala, Sweden).
Immunofluorescence and morphometric analysis
Pancreases from P28 and adult mice were dissected, fixed overnight in 10% formalin neutral buffered solution (Sigma-Aldrich) and embedded in paraffin. Sections (4 μm) were deparaffinised, rehydrated and treated with citrate buffer (10 mmol/l; pH 6.0) in a microwave oven. Primary antibodies used were: rabbit anti-RKIP (1:1,000; Calbiochem, Darmstadt, Germany), guinea pig anti-insulin (1:2,500; Dako, Glostrup, Denmark) and rat anti-cytokeratin 19 (DSHB, Iowa City, IA, USA). Cy2- and Cy3-labelled secondary antibodies (1:200; Jackson ImmunoResearch, Newmarket, UK) were used for immunofluorescence. Hoechst 33258 (Sigma-Aldrich) was used as nuclear marker. Images were taken with a Leica DMR HC epifluorescence microscope and a Leica TCS-SL confocal microscope (Leica Microsystems, Wetzlar, Germany).
For morphometric analysis, serial sections from three different levels for each pancreas block (six sections/animal) were stained with guinea pig anti-insulin antibody as used for immunofluorescence or with mouse anti-glucagon (1:1,000; Dako) using a modified avidin–biotin–peroxidase method and counter-stained with toluidine blue . Slides were analysed using Image J (National Institutes of Health, Bethesda, MD, USA) software.
Proliferation and apoptosis measurements
For proliferation analysis, sections were stained with mouse anti-Ki67 (1:20; Dako, Denmark) as indicated above. A minimum of 3,000 insulin-positive cells per pancreas was counted in adult, P14 and P28 pancreases. For detection of apoptosis, the ApoAlert DNA fragmentation Assay Kit (Clontech, Mountain View, CA, USA) for tissue sections was used following manufacturer’s instructions.
Protein extracts from whole pancreas and islets were prepared in lysis buffer (50 mmol/l Tris pH7.5, 5 mmol/l EDTA, 150 mmol/l NaCl, 1% (vol./vol.) Triton X-100, 10 mmol/l sodium phosphate, 10 mmol/l sodium fluoride, 10 mmol/l sodium orthovanadate and protease inhibitors). Antibodies used for western blot were: RKIP (1:6,000), phosphorylated extracellular signal-regulated kinase 1 or 2 (ERK1/2) and total ERK1/2 (1:1,000; Cell Signaling, Beverly, MA, USA), GSK3 (1:1,000; Santa Cruz Biotechnology, Santa Cruz, CA, USA) and, as loading controls, actin (1:5,000; GE Healthcare, Hertfordshire, UK) or tubulin (1:1,000; Sigma-Aldrich).
RNA isolation and quantitative PCR analysis
Total RNA was prepared from whole pancreas using RNAgents Total RNA Isolation System (Promega, Madison, WI, USA) and from isolated islets using RNeasy Micro Kit (Qiagen, Venlo, the Netherlands). RNA was treated with DNase (Qiagen), and integrity was assessed using the Agilent 2100 Bioanalyzer (Agilent Technologies, Santa Clara, CA, USA). RNA was reverse transcribed using SuperScript VILO (Invitrogen, Carlsbad, CA, USA) following the manufacturer’s instructions. Real-time PCR was carried out in a LightCycler 480 System (Roche) using SYBR Green (Mesa Green, Eurogentec, Seraing, Belgium). The primer sequences used are presented in electronic supplementary material Table 1. Expression levels were normalised to the expression of the gene encoding TATA box binding protein (Tbp) gene and expressed relative to the mean of the values of the same gene found in control wild-type (WT) animals, set at 1.
RNA from three WT and three Rkip1 −/− pancreases was reverse transcribed, labelled and hybridised to mouse genome 430A 2.0 arrays (Affymetrix, Santa Clara, CA, USA) according to the manufacturer’s protocol. Raw data was background-adjusted, normalised and summarised by robust multi-array analysis (RMA) using the Affy package . Raw and processed data were submitted to several quality controls and analysed for statistical differences as described previously . Briefly, analysis was performed with the LIMMA software package available from Bioconductor (www.bioconductor.org) , adjusting p values using Benjamini and Hochberg’s method to control the false discovery rate at a value <5% . Genes were clustered and represented in a heat diagram with dChip software (http://biosun1.harvard.edu/complab/dchip) . Data have been deposited in NCBIs Gene Expression Omnibus (www.ncbi.nlm.nih.gov/geo/), accession number GSE31150.
Data are expressed as mean ± SEM and statistical significance determined by Student’s t test. A p value <0.05 was considered statistically significant.
RKIP1 levels in the mouse pancreas
Glucose homeostasis in adult Rkip1 −/− animals
Beta cell function and mass in Rkip1 −/− animals
No differences in islet cell composition or beta cell volume were apparent between WT and Rkip1 −/− mice (Fig. 4c,d). Yet, Rkip1 −/− pancreases displayed significant alterations in islet size distribution, with more small islets and fewer medium and large islets in Rkip1 −/− compared with WT animals (Fig. 4e). Islet density tended to be higher in Rkip1 −/− mice than in controls (Fig. 4f), which could explain why, despite exhibiting a larger proportion of small islets, the relative beta cell area remained similar in Rkip1 −/− mice to that in WT animals. The percentage of islets located near ducts (<5 μm) was significantly increased in Rkip1 −/− animals (Fig. 4g).
Taken together, these data indicate that an augmented beta cell mass:body weight ratio, rather than alterations in beta cell functionality, is responsible for the improvement in whole body glucose control in Rkip1 −/− mice . Furthermore, while the relevance of altered islet size distribution and location in the control of glucose homeostasis remains unclear, our findings unravel additional potential effects of RKIP1 on islet morphogenesis.
Beta cell proliferation in Rkip1 −/− animals
Our previous results suggested that RKIP1 might be involved in restricting beta cell expansion specifically around the lactation–weaning transition (P21–P28). To further support this possibility, we assessed whether changes in endogenous Rkip1 expression correlated with inhibition of beta cell proliferation during this transition in control mice. We observed a stepwise rise in the abundance of pancreatic Rkip1 mRNA, accompanied by a significant drop in beta cell proliferation rates between P14 and P28 (Fig. 5c,e). Remarkably, increased Rkip1 expression was not correlated with an increase in insulin (Ins1 and Ins2) gene expression (Fig. 5d). Thus, these findings support the notion that activation of Rkip1 expression could participate in restricting beta cell expansion at the end of lactation.
To elucidate the molecular basis for the enhanced beta cell proliferation in P28 Rkip1 −/− animals we examined the phosphorylation status of the MAPKs ERK1/2, which are the canonical targets of the Raf-1/MAP kinase kinase (MEK) pathway. However, we found no significant changes in basal ERK1/2 phosphorylation in islets freshly isolated from Rkip1 −/− mice relative to those from controls (Fig. 5f). RKIP1 also enhances GSK3β-dependent signalling through stabilisation of this kinase . GSK3β is involved in β-catenin turnover and WNT/β-catenin signalling has been shown to regulate neonatal beta cell expansion . Accordingly, GSK3β deficiency results in expanded beta cell mass . These findings prompted us to examine whether the absence of RKIP1 affected GSK3β protein accumulation. However, we found no differences in GSK3β protein or Axin2 mRNA expression (an endogenous indicator of WNT/β-catenin signalling) between P28 Rkip1 −/− and WT islets (ESM Fig. 2). Thus, although alterations in a small subset of beta cells or transient modulation of these pathways cannot be ruled out, these results indicate that neither MAPK nor GSK3β are constitutively changed in Rkip1 −/− islets.
Pancreatic growth in Rkip1 −/− animals
To elucidate the pathways responsible for increased pancreatic growth in Rkip1 −/− mice, we performed whole genome expression profiling using pancreases from P28 Rkip1 −/− and WT animals. Gene ontology analysis identified the cell cycle as the most significantly represented pathway among RKIP1-regulated genes (ESM Fig. 3). Among the genes most highly upregulated in P28 Rkip1 −/− pancreases relative to controls, we identified those encoding the four members of the superfamily of C-type lectins known as regenerating islet-derived 3: Reg3a, Pap (also known as Reg3b), Ingap (also known as Reg3d) and Reg3g. Interestingly, Reg3 genes have been shown to function as stress response genes, antiapoptotic and growth factors in the digestive system and to specifically promote beta cell proliferation in the pancreas . We confirmed these data by quantitative RT-PCR and found that Reg3 overexpression was only seen after weaning (ESM Fig. 3). However, Reg3 mRNAs were not overexpressed in isolated islets from P28 Rkip1 −/− relative to WT animals (MPJ, RG unpublished observations), pointing to the exocrine pancreas as being responsible for Reg3 gene activation in Rkip1 −/− animals.
Taken together, these data reveal RKIP1 as a negative regulator of postnatal pancreatic cell proliferation. Since cell cycle genes are mainly regulated at the post-transcriptional level, the molecular signature of Rkip1 ablation and its significance in pancreatic growth will need to be further explored through the identification of changes in cell cycle protein production in Rkip1 −/− pancreases.
Rkip1 deletion ameliorates STZ-induced diabetes
How beta cell mass is regulated is a critical issue for understanding the pathophysiology of diabetes, which is characterised by a complete (type 1) or relative (type 2) deficiency in beta cell number. During mouse embryogenesis, new beta cells arise through the proliferation and differentiation of a population of progenitor cells. During postnatal development, between birth and weaning, there’s an additional and massive increase in beta cell mass that mainly involves replication from pre-existing beta cells, even though neogenesis is maintained at least during the first week of life. By the time of weaning, beta cell proliferation decreases and is maintained relatively low throughout adulthood (for reviews, see [26, 32]). In this study we aimed to investigate the physiological role of RKIP1 in the regulation of beta cell mass. Our results indicate that RKIP1 participates in the control of pancreatic growth, particularly at the time of weaning. In addition, we demonstrate that the absence of RKIP1 confers significant protection against STZ-induced diabetes, thus indicating that this protein may negatively influence beta cell regeneration after injury in the adult pancreas.
Rkip1 −/− animals exhibit increased beta cell mass relative to body weight and this augmentation, in the absence of changes in insulin sensitivity, is the likely cause for their improved glucose tolerance. Indeed, beta cell mass expansion and body weight follow a linear correlation throughout life , and alterations in this correlation are known to affect glucose homeostasis . A recent study showed that beta cell-specific deletion of Raf-1 kinase results in the downregulation of Ins2 gene transcription and, consequently, in blunted insulin secretion . In contrast, we have found that the insulin content, insulin gene expression and glucose-induced insulin secretion of Rkip1 −/− islets is similar to that of WT islets (Fig. 3 and data not shown), arguing against a major involvement of altered islet function in enhanced whole body glucose control in Rkip1 −/− mutant mice. Yet, it is conceivable that, in adult islets under normal physiological conditions, Raf-1 kinase is already maximally stimulated, so that elimination of RKIP1 has no measurable effects on Raf-1 kinase-dependent processes.
In accordance with prior in vitro experiments , we have observed that RKIP1 inhibits beta cell proliferation in vivo. Remarkably, the inhibitory effect of RKIP1 is only evident during early postnatal stages when the beta cell proliferative capacity is maximal. This age-dependent loss of the effects of RKIP1 may simply reflect changes in the repertoire of signalling factors available to the beta cell at any given moment in time. Alternatively, changes in the responsiveness of beta cells to specific signalling molecules, rather than their availability, may underlie the age-dependent effects of Rkip1 ablation. In line with this, Chen et al  have recently linked age-triggered loss of beta cell expression of the platelet-derived growth factor (PDGF) receptor to reduced beta cell proliferation. Hence, it is tempting to speculate that the proliferative response to Rkip1 ablation will be lost in parallel with the loss of the ability of the beta cell to respond to specific factors.
Genetic mouse models and in vitro studies have helped unravel the identity of growth factors that stimulate beta cell replication, but their role under normal physiological conditions remains uncertain. Among the pathways implicated, many transduce signals through the Raf-1/MEK/ERK cascade, including insulin, insulin-growth factor, hepatic growth factor, epidermal growth factor and PDGF [34–36], and thus could potentially be targeted by RKIP1. Yet, our data indicate that constitutive ERK1/2 activation may not be linked to enhanced beta cell proliferation in Rkip1 mutants. In a similar way, deletion of Raf-1 does not result in decreased ERK1/2 phosphorylation in pancreatic islets, indicating that other Raf-1-dependent pathways may be operative in beta cells . Additionally, RKIP1 may function through Raf-1 independent mechanisms involving NF-κB , GSK3β  or G protein-coupled receptors . Apoptotic rates are comparable between Rkip1 −/− and WT P28 pancreases, implying that NF-κB-dependent survival pathways are not affected by Rkip1 deletion. On the other hand, RKIP1 deficiency does not alter levels of GSK3β in islets, as reported in other tissues , ruling out this possible mechanism. Future work should be aimed at delineating the signal cascades regulated by RKIP1 and their connection to beta cell mitogens during physiological beta cell growth.
This study reveals that RKIP1 is not only involved in restricting beta cell proliferation but it may also inhibit postnatal pancreatic growth. Gene expression profiling has identified the Reg3 gene family as potentially involved in the stimulatory effect of RKIP1 ablation on pancreatic growth at P28. Interestingly, the Reg3 genes have been previously associated to pancreatic regeneration and islet cell neogenesis [37–41]. Further supporting a role for Reg3 genes in pancreatic postnatal development, a recent study has described the upregulation of Reg3 genes in the mouse pancreas between P7 and P30 followed by their subsequent downregulation in adulthood . In this scenario, our data suggest that RKIP1 may participate in restricting Reg3 gene expression after weaning and, in this way, control pancreatic growth. The question of whether exocrine Reg3 gene activation might contribute to enhanced beta cell proliferation in Rkip1 −/− mice remains unresolved. Double Reg3 and Rkip1 mutants and beta cell-specific Rkip1 −/− mice will be necessary to properly address this issue.
In summary, in this paper we show that RKIP1 inactivation results in enhanced beta cell and pancreatic growth in postnatal stages, leading to an increment in beta cell mass and improved glucose tolerance in adult animals. Furthermore, Rkip1 deletion confers recovery to STZ-induced hyperglycaemia in adult animals. These results have therapeutic implications for the development of novel strategies to trigger beta cell mass expansion in diabetes.
The authors thank A. Eriksson (Umeå University) for the optical projection tomography analyses.
This work has been supported by grants from: Spanish Ministerio de Ciencia e Innovación (SAF2010-19527 to R. Gomis and BFU2008-02299/BMC to R. Gasa), Generalitat de Catalunya (2009 SGR 1426 to R. Gomis), Fondecyt no. 1090694 (Chile) and Fundación Marcelino Botin (project 08/351) and the European Community’s Seventh Framework Programme (FP7/2007-2013) under grant agreement no. PIAP-GA-2008-218130 (to R. Gomis). Predoctoral fellowships were provided by Fundación Marcelino Botin (F. N. Pardo), Conicyt-PhD (F. N. Pardo), IDIBAPS (J. Altirriba) and Spanish Ministerio de Economía y Competitividad (M. Pradas-Juni).
Duality of interest
The authors declare that there is no duality of interest associated with this manuscript.
FNP designed, performed the study, analysed and interpreted data and drafted the manuscript. JA, MPJ, AG and UA made substantial contributions to the acquisition, analysis and interpretation of data and critically revised the manuscript. AB, JCS and AJY contributed to data analysis and interpretation and critically revised the manuscript. RGo and RGa designed the study, supervised the project, interpreted data and wrote the manuscript. All authors have approved the final version of the manuscript.